Muhammad-Zain - AI Roast App

Welcome to My Project Page

I'm thrilled to introduce my AI Roast + Judgment App, a React Native project for iOS and Android that delivers humorous, AI-generated roasts and judgments based on user-submitted stories. Users post stories to a feed, vote on others’ submissions, leave comments, and receive witty roasts or verdicts from an AI chatbot powered by advanced language models. The app includes real-time interactions, push notifications, and robust moderation to ensure a safe, fun experience. As a solo developer, I’m targeting a working MVP in 8 weeks, with weekly blog posts to document my progress.

Why I'm Doing This

Creative Tech Challenge: Building an app that combines AI, humor, and social interaction pushes my React Native and backend skills, from real-time feeds to safe AI moderation.
Portfolio Showcase: This project highlights my ability to deliver innovative, user-focused apps, with blog posts offering transparency into my development process.
AI Exploration: I’m experimenting with models like GPT-4 and Grok 3 to craft engaging roasts, with plans to fine-tune open-source models like LLaMA post-launch.
Fast-Paced Goal: I aim to launch a polished MVP by May 31, 2025, with core features like story feeds, voting, and AI roasts, under an 8-week timeline.

What I've Tried So Far

Setting Up the Foundation:
- Initialized React Native with Expo for rapid iOS/Android development.
- Deployed a Node.js + Express server on Heroku, connected to Firebase Firestore for story, vote, and comment storage.
- Blogged about it: “Kicking off AI Roast with Expo and Firebase.”
Story Feed and Voting:
- Built a story submission screen and feed UI, using Firebase for real-time updates.
- Implemented upvote/downvote logic, with Firestore listeners for vote counts.
- Shared a code snippet in my Week 1 post: “Firestore listener for real-time story feeds.”
Comment System (In Progress):
- Adding a comment section for each story, using React Native components.
- Debugging Firestore listener duplication issues to ensure smooth updates.
- Testing on emulators for consistent performance.

Where Things Stand

Core Features Live: Story submission and voting are functional, with real-time feed updates working smoothly.
Comments In Progress: Comment UI and backend are halfway done, resolving Firebase duplication bugs.
Data Ready: Firestore stores stories, votes, and comments, with moderation flags in place.
Blogging on Track: Two posts published (Week 1: Setup, Week 2: Voting), with Week 3 drafting on comment progress.

What's Next

Complete Comments: Finish comment system by Week 3, fixing Firebase listener bugs.
AI Chatbot: Integrate GPT-4 for roasts and judgments in Week 4, testing prompts for humor and safety.
Notifications: Add Firebase FCM for vote/comment alerts in Week 5.
Moderation: Implement AI moderation and GDPR/CCPA consent in Week 6, with a privacy policy by Week 7.
Launch: Polish UI, fix bugs, and submit to App Store/Google Play by Week 8 (May 31, 2025).
Blog Weekly: Share updates with screenshots and snippets (e.g., “GPT-4 roast prompt engineering”).
Post-MVP: Fine-tune LLaMA with Reddit data for sharper roasts.

Future Plans

Advanced AI: Test Grok 3 for spicier roasts or fine-tune Mixtral for custom humor.
Social Features: Add roast cards or verdict meters post-MVP for viral appeal.
Scalability: Move to AWS if user base grows beyond 500.
Legal Compliance: Consult a lawyer for GDPR/CCPA, especially for AI moderation.
Blog Finale: Write “How I Built AI Roast in 8 Weeks,” linking to the live app.

Relevant Links

This project blends AI, humor, and social interaction, showcasing my ability to build engaging apps under tight deadlines. Thank you for exploring my work!

For more details, code samples, and weekly updates, check my blog or connect on LinkedIn !